Content uploaded by Sarah Pogue
Author content
All content in this area was uploaded by Sarah Pogue on May 04, 2021
Content may be subject to copyright.
RESEARCH ARTICLE
Targeting Attenuated Interferon-αto
Myeloma Cells with a CD38 Antibody
Induces Potent Tumor Regression with
Reduced Off-Target Activity
Sarah L. Pogue
1
, Tetsuya Taura
1
, Mingying Bi
1
, Yong Yun
1
, Angela Sho
1
, Glen Mikesell
1
,
Collette Behrens
2
, Maya Sokolovsky
3
, Hussein Hallak
3
, Moti Rosenstock
3
, Eric Sanchez
4
,
Haiming Chen
4
, James Berenson
4
, Anthony Doyle
2
, Steffen Nock
1
, David S. Wilson
1
1Teva Pharmaceuticals, Global Branded Biologics Division, Redwood City, Califor nia, United States of
America, 2Teva Pharmaceuticals, Global Branded Biologics Division, Sydney, Australia, 3Te va
Pharmaceuticals, Global Branded Biologics Division, Netanya, Israel, 4The Institute for Myeloma and Bone
Cancer Research, West Hollywood, California, United States of America
*Sarah.Pogue@tevapharm.com
Abstract
Interferon-Į(IFNĮ) has been prescribed to effectively treat multiple myeloma (MM) and other
malignancies for decades. Its use has waned in recent years, however, due to significant tox-
icity and a narrow therapeutic index (TI). We sought to improve IFNĮ’s TI by, first, attaching it
to an anti-CD38 antibody, thereby directly targeting it to MM cells, and, second, by introducing
an attenuating mutation into the IFNĮpor tion of the fusion protein rendering it relatively inac-
tive on normal, CD38 negative cells. This anti-CD38-IFNĮ(attenuated) immunocytokine, or
CD38-Attenukine™, exhibits 10,000-fold increased specificity for CD38 positive cells in vitro
compared to native IFNĮand, significantly, is ~6,000-fold less toxic to normal bone marrow
cells in vitro than native IFNĮ. Moreover, the attenuating mutation significantly decreases
IFNĮbiomarker activity in cynomolgus macaques indicating that this approach may yield a
better safety profile in humans than native IFNĮor a non-attenuated IFNĮimmunocytokine. In
human xenograft MM tumor models, anti-CD38-IFNĮ(attenuated) exerts potent anti-tumor
activity in mice, inducing complete tumor regression in most cases. Furthermore,anti-CD38-
IFNĮ(attenuated) is more efficacious than standard MM treatments (lenalidomide, bortezo-
mib, dexamethasone) and exhibits strong synergy with lenalidomide and with bortezomib in
xenograft models. Our findings suggest that tumor-targeted attenuated cytokines such as
IFNĮcan promote robust tumor killing while minimizing systemic toxicity.
Introduction
Multiple myeloma (MM) is the second most common blood cell malignancy in the U.S. after
non-Hodgkin’s lymphoma [1,2]. Current treatments for MM include chemotherapy, steroids,
immunomodulatory drugs, proteasome inhibitors and stem cell transplantation. Despite the
PLOS ONE | DOI:10.1371/journal.pone.0162472 September 9, 2016 1/20
a11111
23(1 $&&(66
Citation: Pogue SL, Taura T, Bi M, Yun Y, Sho A,
Mikesell G, et al. (2016) Targeting Attenuated
Interferon-αto Myeloma Cells with a CD38 Antibody
Induces Potent Tumor Regression with Reduced Off-
TargetActivity. PLoS ONE 11(9): e0162472.
doi:10.1371/journal.pone.0162472
Editor: Claire M. Edwards, University of Oxford,
UNITED KINGDOM
Received: March 4, 2016
Accepted: August 23, 2016
Published: September 9, 2016
Copyright: © 2016 Pogue et al. This is an open
access article distributed under the terms of the
Creative Commons Attribution License, which permits
unrestricted use, distribution, and reproduction in any
medium, provided the original author and source are
credited.
Data Availability Statement: All relevant data are
within the paper.
Funding: All studies presented in this manuscript
were funded solely by Teva Pharmaceuticals and was
not by any outside sources. Teva Pharmaceuticals
provided support in the form of salaries for authors
SP, TT, MB, YY, AS, GM, CB, MS, HH, MR, AD, SN,
and DW, but did not have any additional role in the
study design, data collection and analysis, decision to
publish, or preparation of the manuscript. The specific
roles of these authors are articulated in the 'author
contributions' section. Authors JB, HC, and ES are
increased efficacy of these therapies, nearly all patients eventually relapse and become refrac-
tory to treatment [3]. Thus, MM remains an incurable disease with a 47% five-year survival
rate [1,3,4].
IFNċis a pleiotropic proinflammatory cytokine with demonstrated anti-proliferative, cyto-
toxic and anti-neoplastic immunomodulatory activity [5,6]. It has been used for decades to
treat viral infections and certain cancers including MM [7]. While initial trials testing IFNċas
maintenance therapy for MM yielded inconsistent results, subsequent meta-analyses showed
significant improvement in survival rates, although tolerability was poor [8]. The range of seri-
ous side effects frequently associated with IFNċinclude nausea, severe flu-like symptoms, vas-
culopathic complications (e.g., decreased leucoc ytes and platelets), and sometimes depression
or anxiety [9–12]. In one MM study, maintenance therapy with IFNċwas discontinued in up
to 37% of patients in due to toxicity [13]. Such widespread toxicity coupled with the typically
high doses of IFNċrequired for eff icacy in MM patients translates into a narrow therapeutic
index (TI) for IFNċ, defined as the ratio between maximum tolerated dose and minimum ther-
apeutic dose. The narrow TI of IFNċhas limited its consistent clinical use for the treatment of
MM.
One approach to decrease the marked toxicity of cytokines in general in cancer therapy is to
attach them to tumor-targeting antibodies or antibody fragments. This promotes increased
local concentration of the cytokines at tumor sites [14,15]. Such “immunocytokines” have
been describ ed extensively, including those based on IFNċ[16–24]. While potentially decreas-
ing the effective dose, this strategy does not address, and may compound, the issue of IFNċ
toxicity due to the extended half-life generally observed with antibody based therapies and the
ubiquitous expression of the interferon-ċreceptor (IFNAR) on non-tumor cells.
Here, we describe our approach to broaden the TI of IFNċby minimizing its systemic toxic-
ity while retaining its potent anti-tumor activity. We chose the MM tumor antigen CD38 as
our target antigen because it is expressed at high levels on nearly all MM tumor cells and has
limited normal tissue expression [25–27]. We engineered a mutation into the IFNċportion of
the CD38-targeted immunocytokine to significantly reduce its binding to IFNAR on CD38-ne-
gative cells. Our data shows that this CD38-targeted, attenuated IFNċimmunocytokine,
dubbed “CD38-Attenukine™”, is orders of magnitude less potent at stimulating antigen-nega-
tive cells than native IFNċ, and yet maintains potent anti-tumor activity on antigen-positive
cells. In most cases, treatment with CD38-targeted IFNċattenuated Attenukine™ leads to com-
plete elimination of even very large, established human MM tumors in mice.
Materials and Methods
IFNĮconstructs and fusion proteins
Reference anti-CD38 antibody variable regions were generated by PCR from published V
region sequences (reference antibody [28] as described in WO 2013/059885). Negative control,
non-targeted irrelevant specificity, V-region sequences (anti-yellow fever virus clone 2D12
[29]) were generated from published sequences (WO 2013/059885). Negative control
sequences (anti-respiratory syncytial virus) used in the cynomolgus study were generated from
published sequences (WO 2013/059885). The human IFNċ2b gene was isolated from HEK293
genomic DNA by standard PCR methods using primers (5’-GGTAAATCCGGAGGCGGC
GGGAGCTGTGATCTGCCTCAAACCCACAGCCTG-3’ and 5’ACGTGGATCCTATTCCT T
ACTTCTTAAACTTTCTTGC-3’). Attenuating mutations in IFNċ2b were introduced by PCR
at residues which interact with the high affinity IFNċreceptor chain, IFNAR2 [30]. Anti-CD38
or irrelevant V-region and IFNċ2b gene fragments were cloned into the pTT5 mammalian
expression vector [31] containing human IgG4 and kappa immunoglobulin constant region
CD38-Targeted , Attenuated IFNĮIs Potent with Decreased Off-Target Activity
PLOS ONE | DOI:10.1371/journal.pone.0162472 September 9, 2016 2/20
part of a non-profit organization, Institute of Myeloma
and Bone Cancer Research, which performed studies
in this manuscript. Teva Pharmaceuticals funded
these studies.
Competing Interests: This study was funded by
Teva Pharmaceuticals. Sarah L. Pogue, Tetsuya
Taura,Mingying Bi, Yong Yun, Angela Sho, Glen
Mikesell, Collette Behrens, Maya Sokolovsky,
Hussein Hallak, Moti Rosenstock, Anthony Doyle,
Steffen Nock, and David S. Wilson are employed by
Teva Pharmaceuticals. In addition, Sarah Pogue,
David Wilson, TetsuyaTaura, Anthony Doyle, and
Glen Mikesell have filed an International Patent
Application WO 2013/059885 entitled: Polypeptide
Constructs and Uses Thereof. This manuscript
describes a research molecule, CD38-Attenukine™
that is being developed as a Teva product to treat
myeloma. There are no marketed products to
declare. These declarations do not alter our
adherence to all the PLOS ONE policies on sharing
data and materials, as detailed online in the guide for
authors.
genes. Naked antibody and antibody-IFNċfusion proteins were transiently expressed in
HEK293-6E cells [31] and purified using Protein G-Sepharose columns (GE Healthcare, Pis-
cataway, NJ). ELISA binding assay used to determine relative affinities of the wild type and
mutated IFNċfusion proteins for the IFNAR2. Specifically, 96-well plates were coated with
5 μg/ml human IFNAR 2-human Fc fusion protein (R & D Systems, Minneapolis, MN). After
brief washing with PBS, wells were blocked with Superblock blocking solution (ThermoFisher
Scientific, Waltham, MA) and loaded w it h anti-CD38-IFNċfusion proteins diluted in PBS
with 0.05% Tween 20 (PBST). After 60 min incubation at room temperature, wells were
washed with PBST, and anti-human Ĕ-horseradish peroxidase conjugate (SouthernBiotech,
Birmingham, AL) diluted 1/5000 in PBST was applied to each well to capture IFNAR2-anti-
CD38-IFNċfusion complexes. Following 60 minutes incubation and washing with PBST,
3,3’,5,5’-tetramethylbenzidine (TMB, Sigma Aldrich, St. Louis, MO) was added to wells to
develop the colorimetric signal which was measured by a vmax plate reader after color develop-
ment was stopped with sulfuric acid.
Bone marrow colony forming assays
Frozen bone marrow (BM) mononuclear cells from MM patients (AllCells, Inc., Alameda, CA)
and MM cell lines (American Type Culture Collection,ATCC) were cultured in MethoCult
H4230 media (Stem Cell Technologies, Vancouver, Canada) with 10% FBS and 10% phytohe-
magglutin (PHA) stimulated leukocyte conditioned medium. Normal BM cells (AllCells, Inc.)
were cultured in MethoCult H4434 media. All BM cells were incubated with either vehicle
(PBS) or 10,000 IU/ml IFNċIntron1A (Schering Corp. Merck, NJ) at 37°C in 5% CO2 for 7
or 14 days before colonies were stained with SYBR green and visually counted using a micro-
scope. Bone Marrow cells: Normal and myeloma bone marrow cells were purchased from All-
Cells, Inc. as frozen vials containing ten million bone marrow mononuclear cells/vial. All
donors signed an informed consent release and were made aware that the cells were to be used
for research purposes. Donors were all over the age of 18.
MM cell viability and IFNĮactivity assays
Viability of MM cell lines cultured with and without 10,000 IU/ml IFNċ2b (US Biological,
Salem, OR) was determined using the CellTiter-Glo1luminescent assay (Promega, Madison,
WI). MM cell lines included ARH-77 (ATCC1CCL-155™), RPMI8226 (ATCC1CRL-
1621™), NCI-H929 (ATCC1CRL-9068™), U266 (ATCC1TIB-196™) and ARP-1 (Myeloma
Institute, University of Arkansas, Little Rock, AK). IFNċactivity on CD38-negative reporter
cells (iLite™ cell line) was determined using the iLite™ Cell Assay following a modifie d manufac-
turer’s protocol (PBL Assay Science, Piscataway, NJ).
Flow cytometry
BM cells from healthy and MM patients (AllCells, Inc.) were blocked with human IgG, washed
with PBS and stained with 50 μg/ml naked CD38 or CD138 antibodies (eBiosciences, San
Diego, CA). Cells were then incubated with goat anti-mouse IgG, F(ab’)2-PE for 20 minutes on
ice and washed twice. FACS acquisition was done on a BD FacsCalibur™ using CellQuest™
software.
Xenograft Studies
All mouse experiments received approval from the Charles River Labs and Los Angeles Bio-
medical Research Institute Animal Care Committees and were conducted according to the
CD38-Targeted , Attenuated IFNĮIs Potent with Decreased Off-Target Activity
PLOS ONE | DOI:10.1371/journal.pone.0162472 September 9, 2016 3/20
Institutional Animal Care and Use Committee and in adherence to the National Institutes of
Health “Guide for the Care and Use of Laboratory Animals” [32] at Charles River Labs and at Los
Angeles Biomedical Research Institute at Harbor-UCLA, both AAALAC accredited institutions.
Any animal exhibiting weight loss, lethargy, hunched posture, or ruffled f ur that did not improve
after treatment holiday was euthanized. When tumor volume reached endpoint, or at the end of
study, animals were euthanized by the use of CO2 for rodent euthanasia per institutional proto-
cols. In subcutaneous models, mice were euthanized when tumor volume endpoint was reached.
In these studies, no mice exhibited weight loss, lethargy, hunched posture or ruffled fur. In the
systemic myeloma model, MM1S, 30 of 40 mice (all control groups), were euthanized due to
weight loss, lethargy and ruffled f ur (symptoms of disease, study endpoint) prior to end of study.
Non-human Primate study
Source. The cynomolgus monkeys (Macaca fasicularis) used in this study were of Chinese
origin and were received from Primus Bio-Resources Inc., 531, boulevard des Pairies, Bldg 25,
Laval, Quebec, Canada, H7V 1B7 on April 10, 2013. A total of 23 animals (3–4 animals p er test
group) were treated and evaluated.
Animal Care Committee. The non-human primate study was performed by AAALAC
accredited ITR Laboratories Canada, Inc., Baie D'Urfé, Québec, Canada. The study plan was
reviewedand assessedby the Animal Care Committee (ACC) of ITR. ACC acceptance of the
study plan was maintained on file at ITR. All animals used on this study were cared for in
accordance with the principles outlined in the current "Guide to the Care and Use of Experi-
mental Animals" as published by the Canadian Council on Animal Care [33]andthe"Guide
for the Care and Use of Laboratory Animals", a NIH publication [32]. The study did not unnec-
essarily duplicate previous experiments.
Health Status. On arrival at ITR, monkeys were weighed and then subjected to a detailed
physical examination to ensure satisfactory health status. In addition, stool samples were col-
lected and examined for the presenceof fecal parasites using the fecal flotation method [34].
During the quarantine period, all animals were tested twice for tuberculosis by intradermal
injection of tuberculin.
Housing. Each monkey was housed in a stainless steel cage equipped with an automatic
watering system. Each cage was labeled with a color-coded cage card indicating the study,
group and animal numbers, sex and dose level. Animals were housed a total of 15 days individ-
ually. Cages were all in the same room providing animals with visual, olfactory and auditory
stimulation/access to one another.
Room Environment. The animal room environment was controlled (targeted ranges:
temperature 21 ± 3°C, relative humidity 50 ± 20%, 12 hours light, 12 hours dark [except during
designated procedures] and a minimum of 10 air changes per hour). Temperature and relative
humidity were monitored continuously and records are maintained at ITR.
Diet/Water. A standard certified commercial primate chow (Teklad Cer tified Primate
Chow #2055C) was available ad libitum to each monkey except during designated procedures.
Municipal tap water (which was purified by reverse osmosis, ultraviolet light and further fil-
tered with a 0.2 μm filter) was provided to the animals ad libitum except during designated
procedures.
Environmental Enrichment. Animals were offered certified treats, non-certified treats
(e.g., frozen yogurt, banana, or grapes; nuts, peanuts, raisins) and other non-dietary items (e.g.,
toys) as part of the ITR environmental enrichment program at appropriate intervals. More spe-
cifically, for appetite enhancement, fruit, vegetables or certified/non-certifiedtreats were
offered to each monkey once daily during all phases of the study.
CD38-Targeted , Attenuated IFNĮIs Potent with Decreased Off-Target Activity
PLOS ONE | DOI:10.1371/journal.pone.0162472 September 9, 2016 4/20
Acclimation. An acclimation period of approximately 4 weeks (approximately 3 weeks of
quarantine and 1 week of pretreatment procedures) was allowed between receipt of the animals
and the start of dosing to accustom the monkeys to the laboratory environment. Monkeys were
acclimated to the experimental procedures (e.g., sling restraint for intravenous infusion) for 3
consecutive days prior to the start of dosing.
Administration of the Test Items. Each test item dose formulation was administered
once by intravenous infusion over a period of 1 hour via a disposable indwelling catheter
inserted into one of the saphenous veins.
Monitoring. All monkeys were monitored daily for mortality, or clinical signs involving
general appearance, posture, gait, respiration, behavioral and coordination abnormalities, and
abnormal feces. More careful examination of an individual animal’s body, mouth, nose, ears,
eyes, skin, and coat were included if any initial abnormalities were noted. Body weights were
monitored weekly as well as on days -2, 1 and 8. No animals became ill or injured at any point
during the study.
Sampling. A series of 10 blood samples (approximately 0.5 mL each for time point up to
48 h post end of infusion and approximately 1 mL each for 96 h to 168 h post end of infusion)
was collected from each monkey at the following time points: Pre-dose, 0 minutes (immedi-
ately post end of infusion), 2, 6, 12, 24, 48, 96, 120 and 168 hours post end of infusion. Serum
samples were evaluated for biomarker analysis.
Terminal Procedure. Following collectionof the last blood samples on Day 8, all animals
were released to the ITR spare colony.
Xenograph tumor models
LAGĔ-1A, LAGĔ-2, and LAGĕ1 tumor fragments (20–40 mm
3
), obtained from in vivo pas-
saged MM patient cells from 3 donors, were implanted intramuscularly into CB.17 SCID
mice as describ ed [35,36]. Treatment started on day 8 following implantation. Tumor vol-
umes, M protein and mouse body weights were measured twice weekly by caliper;endpoint
was reached at tumor volume of 2000 mm
3
. The sources of these tumor cells were: LAGk-2,
date of collection: 11/9/06, 68 year old male, diagnosed with myeloma, bone marrow aspirate
obtained from patient by informed consent, LAGk-1A, date of collection 9/15/04, 76 year old
female, diagnosed with myeloma, bone marrow aspirate obtained by informed consent,
LAGĕ-1, date of collection >10 years ago, female diagnosed with myeloma, tumor cells
obtained by informed consent. Cell line models: 10 million NCI-H929 (ATCC1CRL-
9068™) or Daudi cells (ATCC1CCL-213™) were mixed 1:1 with Matrigel and injected sub-
cutaneously into the hind flank. Tumor volumes and mouse body weights were measured
twice weekly by caliper. As a direct measurement of tumor growth, calipers were used to
assess tumor volume twice weekly, and the formula for an ellipsoid volume was applied 4/3π
x [width/2]
2
x [length/2]. Endpoint was reached at tumor volume of 2000 mm
3
.Forthesys-
temic multiple myeloma model, MM1S 10 million cells (ATCC1CRL-2974™) were injected
intravenously into the tail vein; mouse body weight and overall health were monitored twice
weekly with a survival endpoint.
Statistical Analysis
For all in vitro experiments, means of two or three samples +/- Standard Deviation are pre-
sented. For in vivo studies, tumor volumes are largely represented at mean tumor volume of 10
mice +/- Standard Error of Mean. P values were calculated using Students t-Test.
CD38-Targeted , Attenuated IFNĮIs Potent with Decreased Off-Target Activity
PLOS ONE | DOI:10.1371/journal.pone.0162472 September 9, 2016 5/20
Results
IFNĮinhibits growth of human MM cells
IFNċis known to inhibit tumor growth of numerous types of cancers including lymphomas
and leukemias [37]. To evaluate the effects of IFNċon human MM cells in vitro, we assessed
its activity in colony forming assays using BM samples from MM patients and MM cell lines.
We found that IFNċstrongly inhibited the growth of all tested MM specimens (Fig 1A and
1B). When grown in the presence of 10,000 IU/ml of IFNċcolony counts from primary MM
cells and MM cell lines were reduced by 72–100% compared to untreated controls (Fig 1A). In
cell culture proliferation assays, IFNċtreatment reduced proliferation between 50 to 85% com-
pared to controls (Fig 1B).
To evaluate the activity of IFNċon primary MM tumors in vivo, we performed xenograft
models using in vivo passaged MM tumor fragments [35,36]. Fragments of tumors LAGĔ-1A,
LAGĔ-2 and LAGĕ-1 were implanted intramuscularly into SCID mice on day 0, and IFNċor
Fig 1. High IFNĮexposure inhibits human MM tumor growth in vitro and in vivo. (A) Bone marrow (BM)
cells from 13 MM patients and 2 human MM cell lines were grown in colony supporting matrix with or without
IFNĮ(10,000 IU/ml). Percent reduction in colony number from each donor was calculated relative to
untreated control wells. Representative images for one primar yMM bone marrow sample (#6) are shown with
and without IFNĮtreatment. (B) Five MM cell lines were cultured for 3 days in standard cell media with or
without 10,000 IU/ml IFNĮ. Percent viability of each cell line relative to untreated cells (dotted line) is shown.
Data represent means from triplicate wells +/- SD. (C, D, E) IFNĮtreatment of 6–8 week old CB17 SCID mice
implanted with mouse-passaged human MM tumor fragments (LAGț-1A, LAGț-2, and LAGȜ-1). On day 8
post-tumor implantation, recipient mice (n = 10 per group) were treated with IFNĮor vehicle twice a week for 4
weeks. The treatment windows for all experiments are indicated by shaded bars. Mean tumor volumes +/-
SEM are shown. (C) In a dose-response study, only high doses of IFNĮ(0.12mg/kg and 0.48 mg/kg) were
effective at completely inhibiting LAGț-1A tumor growth. (D) Treatment with IFNĮ0.48 mg/kg) also strongly
inhibited LAGț-2 tumor growth. On Day 56, p = 0.0001. (E) Treatment with IFNĮ(0.24 mg/kg) did not
significantly inhibit LAGȜ-1 tumors (p = 0.7537), which originated from a bortezomib-refractor y MM patient. All
statistical analyses were performed using Student’s t-Test.
doi:10.1371/journal.pone.0162472.g001
CD38-Targeted , Attenuated IFNĮIs Potent with Decreased Off-Target Activity
PLOS ONE | DOI:10.1371/journal.pone.0162472 September 9, 2016 6/20
vehicle treatments were initiated on day 8. In an IFNċdose response study using the LAGĔ-1A
model, human IFNċat doses of 0.12–0.48 mg/kg strongly inhibited growth of tumors in all
(10/10) mice (Fig 1C). Similarly, growth of LAGĔ-2 tumors was also strongly inhibited by high
dose (0.48 mg/kg) IFNċtreatment (Fig 1D). All mice (10/10) were tumor free within two
weeks of treatment and remained so following cessation of treatment. In contrast, LAGĕ-1
tumors, originating from a bortezomib-refractory MM patient, did not show a significant
reduction in tumor growth even with high dose (0.24 mg/kg) IFNċtreatment (Fig 1E). We
note that the 0.12 to 0.48 mg/kg doses of IFNċare orders of magnitude higher than typical
IFNċdoses given to cancer patients (approximately 1μg/kg). Such high IFNċdoses are toler-
ated in animal experiments without toxicity because human IFNċdoes not efficiently stimulate
murine IFNAR [38,39]. These experiments reveal the potential of IFNċto exert profound
direct anti-MM tumor activity in vivo if sufficient exposure is achieved.
Antibody-IFNĮfusion protein displays high specificity for CD38 positive
cells
Despite the strong potential of IFNċto treat MM, its side effects limit clinical use [9–12]. In an
effort to broaden the TI of IFNċ, we first employed a conventional immunocytokine approach.
We produced a fusion protein, designated anti-CD38-IFNċ(wt), consisting of wild-type
human IFNċ2b fused directly to the C-terminus of a human IgG4 anti-CD38 heavy chain (Fig
2A). CD38 is highly expressed on nearly all MM tumor cells and normal human plasma cells,
with low expression on other normal BM precursor subsets (Fig 2B)[40].
When evaluated in vitro, the anti-CD38-IFNċ(wt) immunocytokine proved hig hly effective
at inhibiting CD38 positive ARP1 MM cell proliferation, with a similar potency as that of
unmodified, wild type IFNċ(hereafter referred to as native IFNċ) (IC50s of 3.39 pM vs 4.92
pM, respec tively; Fig 3A). Given the high affinity of IFNċfor its receptor, this fusion protein
also stimulated potent IFNċactivity on CD38 negative cells (Fig 3B) although with measurably
reduced activity compared to native IFNċ(19.5 pM vs 0.726 pM, respectively). This reduced
activity is likely due to steric constraints created by fusion of IFNċto the antibody.
To eliminate the IFNċactivity on normal cells, we engineered various point mutations into
the IFNċportion of the fusion protein to reduce its affinity for IFNAR. We hypothesized that
normal cells which express low or no CD38 would be relatively unresponsive to the attenuated
IFNċ(Fig 2A, lower diagram). The CD38 antibody portion of the immunocytokine would,
however, direct a high local concentration of attenuated IFNċto the surface of CD38 positive
tumor cells (Fig 2A, upper diagram), thereby compensating for the weakened binding of IFNċ
to IFNAR and restoring activity on the targeted tumor cells.
To identify the optimal IFNċmutation, we constructed a large panel of anti-CD38-IFNċ
fusion proteins (Attenukines™) with specific IFNċmutations at residues which interact with the
high affinity IFNċreceptor chain, IFNAR2 [30]. Representative constructs and their relative
binding affinities to IFNAR2 are shown in Ta ble 1 . Each construct was evaluated in vitro for
potency in the CD38 positive and CD38 negative cell line assays. One Attenukine™, designated
anti-CD38-IFNċ(A145D), hereafter referred to as anti-CD38-IFNċ(att), was selected for further
study based on optimal in vitro on-target and off-targetcell activity and high production levels.
Anti-CD38-IFNċ(att) strongly inhibited MM tumor cell proliferation, although it displayed
approximately 20x lower potency relative to native IFNċ(IC50s of 95.3 pM and 4.97 pM, respec-
tively; Fig 3C). More notably, on CD38-negative cells, the anti-CD38-IFNċ(att) Attenukine™ was
200,000-fold less potent than native IFNċ(EC50's of 52,200 pM and 0.260 pM, respectively) (Fig
3D). Thus, anti-CD38-IFNċ(att) exhibited negligible IFNċactivity on normal, CD38 negative
cells while retaining strong anti-tumor activity on CD38 positive tumor cells.
CD38-Targeted , Attenuated IFNĮIs Potent with Decreased Off-Target Activity
PLOS ONE | DOI:10.1371/journal.pone.0162472 September 9, 2016 7/20
To estimate the potential TI increase of anti-CD38-IFNċ(att) and other IFNċAttenukines™
in vivo, we calculated an in vitro “antigen specificity index (ASI)” for each construct (Table 2).
ASI is defined as the relative potency of an Attenukine™ on antigen-positive cells versus anti-
gen-negative cells compared to native IFNċ. An ASI of 1 indicates no antigen specif icity and is
the value assigned to native IFNċ. An ASI of 100 indicates a 100-fold increased potency on
antigen-positive cells (tumor) compared to antigen-negative cells (normal) relative to native
IFNċ.AsshowninTabl e 2 , the ASI for the Attenukine™ anti-CD38-IFNċ(att) is 10,100, repre-
senting approximately 10,000-fold increased specif icity compared to native IFNċfor activity
on CD38 positive versus CD38 negative cells. In contrast, the ASI for the targeted immunocy-
tokine with native IFNċsequence [anti-CD38-IFNċ(wt)], such as those recently described by
others [16–24], is a mere 39. These in vitro findings predict a markedly broadened TI for the
Attenukine™ anti-CD38-IFNċ(att) compared to a non-attenuated IFNċimmunocytokine.
Fig 2. Experimental strategy to enhance the therapeutic index of IFNĮin MM treatment. (A)
Experimental strategy: Left of dotted line, the standard immunocytokine approach targeting CD38 positive
tumor cells (upper scheme) directs wild type IFNĮto the tumor via an anti-CD38 antibody, causing IFNĮ
signaling through IFNAR. Because IFNĮhas such high affinity for its receptor, signalingalso occurs in CD38
negative (normal) cells (lower scheme) leading to systemic toxicity that is typically associated with IFNĮ
treatments. With the Attenukine™approach (rightside of dotted line), the IFNĮportion of the molecule is
mutated to significantly reducebinding to its receptor. Despite the attenuation, anti-CD38-IFNĮ(attenuated)
maintains potent activity on CD38 positive MM cells which is mediated by the strong antibody-antigen
interaction on tumor cells (upperscheme), with little to no effect on CD38 negative normal cells (lower
scheme). Thus, the Attenukine™ approach reduces toxicity on normal cells while maintaining high potency
on antigen positive tumor cells. (B) A FACS plot of normal BM (upper plot) shows that high CD38 expression
is restricted to plasma cells, with low levels of CD38 expressed on a subset of CD138 negative progenitor
cells and lymphocytes. The FACS plot of MM BM (lower plot) shows that MM blasts express high levels of
CD38.
doi:10.1371/journal.pone.0162472.g002
CD38-Targeted , Attenuated IFNĮIs Potent with Decreased Off-Target Activity
PLOS ONE | DOI:10.1371/journal.pone.0162472 September 9, 2016 8/20
Fig 3. Anti-CD38-IFNĮ(att) retains strong on-target tumor cell activity with reduced off-target activity
in vitro. Two assays were used to evaluate the antigen specificity of anti-CD38 targeted IFNĮ’s. The graphs
on the left show viability assays of treated ARP-1 MM cells three days after incubation with IFNĮcompounds.
This assay predicts on-target, anti-tumor activity of IFNĮand IFNĮ-fusion proteins. The graphs on the right
show IFNĮresponses of CD38-negative reporter cells that express a reporter gene driven by the IFNĮ
response element (IRE). This response predicts IFNĮactivity on CD38-negative, normal cells. Both the cell
viability and IRE activity are measured by luminescence and presented as relative luminescence units (RLU).
Data represent means from triplicate measurements +/- SD. (A) Activity of the wild type IFNĮ
immunocytokine, anti-CD38-IFNĮ(wt) (black diamonds), has similar potency as native IFNĮ(gray circles) on
CD38-positive MM ARP1 cell viability. (B) The IRE activity of the wild type IFNĮimmunocytokine is 27-fold
less potent than native IFNĮon the CD38-negative reporter cell line. (C) Theactivity of the attenuated IFNĮ
fusion protein, anti-CD38-IFNĮ(att),is 19x less potent than native IFNĮon the CD38 expressing ARP-1 cell
line. (D) The activity of anti-CD38-IFNĮ(att) is 200,000x weaker than native IFNĮon the CD38-negative
reporter cell line.
doi:10.1371/journal.pone.0162472.g003
Table 1. IFNĮattenuating mutations in CD38- Attenukines™ and relative binding affinitiesto IFNAR2.
Construct Codon and Position Relative Binding Affinity
144 145
ĮCD38-IFNĮ(wt) AGA GCA 1.0
ĮCD38-IFNĮ(A145G) AGA GGA 11.4
ĮCD38-IFNĮ(R144A) GCA GCA 18.6
ĮCD38-IFNĮ(A145H) AGA CAC 34.7
ĮCD38-IFNĮ(A145D)*AGA GAT 55.3
ĮCD38-IFNĮ(R144T) ACC GCA 91.6
ĮCD38-IFNĮ(R144I) ATC GCA 128.0
Italic text indicates mutated codons.
The 144 and 145 residues within the IFNĮmolecule are among those which interactwith the high affinity
IFNĮreceptor chain, IFNAR2.[30]
*The ĮCD38-IFNĮ(A145D) construct is designated as anti-CD38-IFNĮ(att) in text.
doi:10.1371/journal.pone.0162472.t001
CD38-Targeted , Attenuated IFNĮIs Potent with Decreased Off-Target Activity
PLOS ONE | DOI:10.1371/journal.pone.0162472 September 9, 2016 9/20
Reduced activity of anti-CD38-IFNĮ(att) on normal cells
The subsets of hematopoietic cells in normal BM and peripheral blood that express CD38 (Fig
2B) represent the normal cells most likely to be adversely affected by CD38-targeted therapies.
We found that CD38 expression on peripheral and BM lymphocytes is approximately 100-fold
less than that of plasma cells and myeloma blasts (Figs 2B and 4A). Thus, IFNAR stimulation
of lymphocytes in patients by a CD38 targeted attenuated cytokine would be predicted to be
relatively low compared to MM cells. Nevertheless, to evaluate in vitro whether anti-
CD38-IFNċ(att) mediates a toxic response in normal BM, we compared the activity of anti-
CD38-IFNċ(att) versus native IFNċin hematopoietic colony formation assays using cells iso-
lated from healthy human BM donors. We also included an human IgG4 negative control
fusion protein, designated “non-targeted-IFNċ(att)”, that had irrelevant (non-CD38 binding)
variable domains.
The results of these experiments showed that, as expected, native IFNċsignificantly inhib-
ited BM colony formation, yielding, at the highest concentration tested (25 nM), only 25% of
colonies formed compared to untreated cells (Fig 4B). In contrast, BM cultures treated with the
same or 10X higher concentrations of either anti-CD38-IFNċ(att) or a non-targeted-IFNċ(att)
fusion protein maintained good colony formation, yielding approximately 85% of the number
of colonies formed compared to untreated cells. At a much higher concentration (6000 nM),
colony formation dropped to 60% of control cultures for both molecules, corresponding to the
inhibitory effect induced by approximately 1nM native IFNċ. This indicates that attenuated
IFNċ(either targeted to CD38 or not) had approximately 6,000x less anti-proliferative activity
than native IFNċon normal BM cells. Based on these findings, we would predict that the low
level of CD38 expression on normal BM cells and peripheral lymphocytes is not sufficient to
allow anti-CD38-IFNċ(att) stimulation through the IFNċreceptor at relevant drug concentra-
tions and that BM toxicity in vivo may therefore be largely avoided.
To assess decreased off-target activity of the attenuated IFNċantibody fusion in vivo, we
tested the activity of non-targeted-IFNċ(att) versus non-targeted-IFNċ(wt) in cy nomolgus
monkeys which, unlike mice, exhibit relevant and measurable responses to native human IFNċ
Table 2. Antigen specificity index (ASI) of anti-CD38-IFNĮand attenuated variants.
Construct CD38 Positive Cells CD38 Negative Cells
(A) (B) (A/B)
IC50 (pM) IC50 IFN/IC50 variant EC50 (pM) EC50 IFN/EC50 variant Antigen Specificity Index
Exp I
Native IFNĮ4.92 1.00 0.726 1.00 1
ĮCD38-IFNĮ(wt) 3.39 1.45 19.47 3.7x 10
−2
39
Exp II
Native IFNĮ4.97 1.00 0.26 1.00 1
ĮCD38-IFNĮ(A145D)*95.30 0.05 52 500 4.94x 10
−6
10 115
ĮCD38-IFNĮ(A145G) 15.30 0.25 2 040 1.27x 10
−4
1 962
ĮCD38-IFNĮ(A145H) 41.06 0.09 24 900 1.04x 10
−5
8 619
ĮCD38-IFNĮ(R144A) 138.30 0.03 25 800 1.01x 10
−5
2 977
ĮCD38-IFNĮ(R144I) 136.10 0.03 159 000 1.64x 10
−6
18 346
ĮCD38-IFNĮ(R144T) 281.20 0.01 30 800 8.44x 10
−6
1 185
The values for native IFNĮserve as internal reference points within each experiment and as controls for inter-experimental variability between Experiment 1
and Experiment 2.
*The ĮCD38-IFNĮ(A145D) construct is designated as anti-CD38-IFNĮ(att) in text.
doi:10.1371/journal.pone.0162472.t002
CD38-Targeted , Attenuated IFNĮIs Potent with Decreased Off-Target Activity
PLOS ONE | DOI:10.1371/journal.pone.0162472 September 9, 2016 10 / 20
[41]. This study addressed whether normal, antigen-negative cells would be stimulated in vivo
by an attenuated IFNċ-antibody fusion and whether the A145D mutation in IFNċis sufficient
Fig 4. Attenuated IFNĮfusion proteins have minimal activity on normal cells in vitro and in vivo. (A)
CD38 expression on subsets of normal bone marrow cells as measured by FACS. Data represent means of
two donor samples +/- SD. (B) Inhibition of normal bone marrow colony formationby anti-CD38-IFNĮ(att)
(black solid line) or non-targeted-IFNĮ(att) (gray solid line) compared to native IFNĮ(black dashed line). The
graph shows percentage of colony counts relative to untreated, control BM cells. Data represent means of
three donor samples +/- SD. (C) Serum levels of IFNĮinduced biomarkers neopterin and IP-10 in
cynomolgus monkeys (n = 3or 4 per group)after a singledose of non-targeted,attenuated IFNĮfusion
protein (non-targeted-IFNĮ(att)), non-targeted wild type IFNĮfusion protein (non-targeted-IFNĮ(wt)) or 10mg/
kg naked non-targeted control IgG4 antibody. ND = Not Done. Data represent means from three or four
monkey samples +/- SD.
doi:10.1371/journal.pone.0162472.g004
CD38-Targeted , Attenuated IFNĮIs Potent with Decreased Off-Target Activity
PLOS ONE | DOI:10.1371/journal.pone.0162472 September 9, 2016 11 / 20
to render the fusion protein less active than a wild type IFNċfusion. Animals were treated with
a 1-hr intravenous infusion of various doses of either non-targeted-IFNċ(att) or non-targeted-
IFNċ(wt) fusion proteins. Serum levels of two known IFNċ-induced biomarkers, neopterin
and IP-10 [42,43], were measured in monkey serum after each dose as indicators of systemic
IFNċactivity.
Non-targeted-IFNċ(wt) induced substantially higher serum levels of both biomarkers com-
pared to non-targeted-IFNċ(att) (Fig 4C). For example, the levels of both neopterin and IP-10
in the monkeys given 0.1 mg/kg non-targeted-IFNċ(wt) were greater than t he serum levels
induced in monkeys given 10 mg/kg of non-targeted-IFNċ(att). This dosing difference indi-
cates that the attenuated IFNċantibody fusion is at least 100-fold less potent in stimulating
IFNċsignaling on normal, antigen negative tissues than the wild type IFNċantibody fusion
protein in vivo. These in vivo data confirm that the diminished off-target activity of attenuated
IFNċobserved in BM cells in vitro translates to a relevant primate system and suggests that the
tolerable dose of an attenuated IFNċantibody fusion may likely be higher than that of wild-
type IFNċimmunocytokines in patients.
Anti-CD38-IFNĮ(att) induces potent tumor regression and increases
survival in xenograft models
To confirm that anti-CD38-IFNċ(att) retains robust anti-tumor activity in vivo despite its
attenuating mutation, we tested its activity in CD38 positive human MM and lymphoma
xenograft tumor models. These included subcutaneous implantation models for myeloma
(NCI-H929 cell line) and lymphoma (Daudi cell line), as well as a systemic model of mye-
loma (MM1S cell line). Treatment with anti-CD38-IFNċ(att) resulted in complete regres-
sion of well-established NCI-H929 MM subcutaneous tumors (Fig 5A). Remarkably, 10 of
10 mice were tumor-free by day 22 of treatment and showed no signs of tumor regrowth for
the duration of the experiment (up to 72 days after treatment cessation). In contrast, all
mice treated with vehicle, anti-CD38 naked antibody, or native IFNċshowed rapid
NCI-H929 tumor growth, reaching tumor endpoint before day 40 (Fig 5A). Anti-
CD38-IFNċ(att) also had strong activity against Daudi lymphoma subcutaneous tumors rel-
ative to other treatments (Fig 5B). All Daudi tumors showed complete regression within 24
days of treatment, although small tumors did recur in 4/10 mice several weeks after treat-
ment cessation in this model. In our studies, native IFNċtreatment was ineffective in tumor
cell models based on subcutaneous xenograft implants (Fig 5), whereas, native IFNċshowed
strong anti-tumor activity in the two LAG xenograft models which were based on in vivo
passaged MM tumor fragments implanted intramuscularly (Fig 1). We postulate that IFNċ
was effective in the latter cases because tumor exposure to IFNċis likely much higher in
tumor fragments growing in hig hly vascularized muscle tissue compared to cells implanted
subcutaneously.
In the systemic MM model using MM1S cells, mouse body weights and overall health were
monitored after the start of treatments, and survival time was the experimental endpoint. As
shown in Fig 5C, all 10 mice treated with anti-CD38-IFNċ(att) survived the full duration of
study (>100 days) compared to a median survival time of 56 days for vehicle treated mice.
Native IFNċincreased median survival time by 18 days compared to vehicle (74 days vs 56
days, p = <0.001), and naked anti-CD38 antibody provided a 5 day median survival benefit
over vehicle (62 days vs 56 days, p = 0.004). All statistical analyses were performed using Stu-
dent’s t-Test. The results from all three xenograft models demonstrated that a strongly attenu-
ated IFNċtargeted directly to MM tumors via a CD38 antibody can induce improved, potent
and long-lasting anti-tumor activity in vivo.
CD38-Targeted , Attenuated IFNĮIs Potent with Decreased Off-Target Activity
PLOS ONE | DOI:10.1371/journal.pone.0162472 September 9, 2016 12 / 20
CD38-Targeted , Attenuated IFNĮIs Potent with Decreased Off-Target Activity
PLOS ONE | DOI:10.1371/journal.pone.0162472 September 9, 2016 13 / 20
Anti-CD38-IFNĮ(att) eliminates very large MM tumors in mice
The robust, and in most cases curative, anti-tumor activity observed in vivo from anti-
CD38-IFNċ(att) therapy prompted further evaluation of this compound in mice bearing larger
tumors. We assessed its activity on NCI-H929 subcutaneous tumors that had reached an aver-
age volume of 730 mm
3
(Fig 6). One mouse, with a tumor volume of 1800mm
3
on day 1 of
treatment, reached endpoint the following day and was excluded from the study. Of the
remaining 8 treated mice, 7 were completely tumor free within 18 days of the start of treatment,
and the remaining mouse was tumor free within 30 days (Fig 6A and 6B). Furthermore, no
tumor regrowth was observedin any of the mice by study endpoint on day 76. These data dem-
onstrate that targeted delivery of an attenuated IFNċmolecule has profound anti-tumor activ-
ity, even on very large tumors.
Fig 5. Anti-CD38-IFNĮ(att) induces significant, durable MM tumor regression and increases survival
time in mice. (A) Inhibition of subcutaneous MM NCI-H929 tumor growth in 8–12 week old CB.17 SCID mice
(n = 10 per group) that were treated twice weekly for 4 weeks with anti-CD38-IFNĮ(att)or naked anti-CD38
antibody (both at 10 mg/kg), native IFNĮ(0.4 mg/kg), or vehicle. Mice were injected subcutaneously with
1x10
7
NCI-H929 myeloma tumor cells in 50% Matrigeland treatments began when tumors reached 120–150
mm
3
. (B) Inhibition of subcutaneous Daudi tumor growth in 6–8 week old irradiated NOD-SCID mice (n = 10
per group) treated twice weekly for 4 weeks with the same test compounds and regimen as in A. Mice were
injected subcutaneously with 1 x 10
7
Daudi Burkitt’s lymphoma tumor cells in 50% Matrigel one day after
irradiation with 200rad (
60
Co). Treatments began after mean tumor size reached 169 mm
3
. (C) Percent
survival of 6–8 week old irradiated NOD-SCIDmice (n = 10 per group) implanted by tail vein injection of 1 x
10
7
MM1S myeloma tumor cells. Seven days later, treatments were given twice weekly for 67 days with the
same test compounds as in A. The treatment windows for all experiments are indicated by shaded bars.
doi:10.1371/journal.pone.0162472.g005
Fig 6. Anti-CD38-IFNĮ(att) induces profound anti-tumor activity on large, well-established NCI-H929
MM tumors in vivo. (A) Regression of very large (median volume = 730 mm
3
) subcutaneous NCI-H929 MM
tumors in 8–12 week old CB.17 SCID mice (n = 8 per group) that were treated twice weekly for 3 weeks with
10 mg/kg of anti-CD38-IFNĮ(att) or vehicle. (B) Tumor volumes of the individual treated mice (n = 8) from the
same experiment. (C) Images of a representative mouse from the same study with an established
subcutaneous NCI-H929 tumor pre- and post-treatment with anti-CD38-IFNĮ(att).
doi:10.1371/journal.pone.0162472.g006
CD38-Targeted , Attenuated IFNĮIs Potent with Decreased Off-Target Activity
PLOS ONE | DOI:10.1371/journal.pone.0162472 September 9, 2016 14 / 20
Anti-CD38-IFNĮ(att) provides greater anti-tumor activity than standard
MM treatments
To assess the potency of anti-CD38-IFNċ(att) compared to standard MM therapeutics, we
tested anti-CD38-IFNċ(att) along with several registered compounds in the NCI-H929 xeno-
graft model. Dexamethasone, lenalidomide, and bortezomib were delivered as single agents
using standard dosing regimens [44–46]. Treatment with anti-CD38-IFNċ(att) generated the
strongest anti-tumor response of all compounds tested in this model (Fig 7A). By day 25 of
treatment, NCI-H929 tumors had completely disappeared in all mice treated with anti-
CD38-IFNċ(att), and all mice survived to the end of the study (day 80). In contrast, lenalido-
mide and bortezomib generated moderate tumor inhibition during treatments (median time to
endpoint [TTE] of 55 days and 43 days, respectively), but tumors regrew in all of these animals
once treatments stopped at 21 days and 28 days, respectively. Dexamethasoneand the naked
anti-CD38 control antibody exerted only a slight delay in NCI-H929 tumor growth compared
to vehicle treatment (TTE 31 days, 26 days, and 18 days, respectively).
In follow up experiments, lenalidomide (25 mg/kg) and bortezomib (1mg/kg) were each
delivered in combination with a suboptimal dose (2.5 mg/kg) of anti-CD38-IFNċ(att). Both
combination therapies resulted in a synergistic response with complete tumor regression in all
Fig 7. Enhanced tumor regression and synergistic activity of anti-CD38-IFNĮ(att) compared to
standard MM therapeutics. (A) Tumor volumes from NCI-H929 tumor bearing SCID mice treated with 5mg/
kg anti-CD38-IFNĮ(att), twice a week for 3 weeks or standard MM therapeutic agents including
dexamethasone (10 mg/kg, daily for 28 days), lenalidomide (25 mg/kg, daily for 21 days), bortezomib (1 mg/
kg, twice a week for 28 days), or vehicle. Average tumor size at star t of treatment was 125 mm
3
. (B)
NCI-H929 tumor growth inhibition and regression in SCID mice treated twice a week for 3 weeks with
lenalidomide (25 mg/kg), a sub-optimal dose of anti-CD38-IFNĮ(att) (2.5 mg/kg), a combination of
lenalidomide and anti-CD38-IFNĮ(att),or vehicle. Average tumor size at star t of treatment was 250 mm
3
. (C)
NCI-H929 tumor growth inhibition and regression in SCID mice treated twice a week for 3 weeks with
bortezomib (1.0 mg/kg), a sub-optimal dose of anti-CD38-IFNĮ(att) (2.5 mg/kg), a combination of bortezomib
and anti-CD38-IFNĮ(att), or vehicle. Average tumor size at start of treatmentwas 250 mm
3
. (D) Mouse
passaged human MM xenograph LAGȜ-1 tumor inhibition in SCID mice treated twice a week for 4 weeks with
bortezomib (1.0 mg/kg), a high dose (10.0 mg/kg) of anti-CD38-IFNĮ(att), a combination of bortezomib and
anti-CD38-IFNĮ(att), or vehicle. Treatment star ted on day 8 following implantation.
doi:10.1371/journal.pone.0162472.g007
CD38-Targeted , Attenuated IFNĮIs Potent with Decreased Off-Target Activity
PLOS ONE | DOI:10.1371/journal.pone.0162472 September 9, 2016 15 / 20
mice within two weeks of treatment (Fig 7B and 7C) and no tumor regrowth after treatments
were stopped. We also tested bortezomib plus anti-CD38-IFNċ(att) combined treatment in the
bortezomib-refractoryLAGĕ-1] MM model (Fig 7D). Because mice bearing LAGĕ-1 tumors
were not responsive to treatment with native IFNċ(Fig 1E), we used a higher dose (10 mg/kg)
of anti-CD38-IFNċ(att) combined with the standard dose (1 mg/kg) of bortezomib adminis-
tered twice weekly. This combination proved highly effective in reducing LAGĕ-1 tumor size
in all mice compared to either treatment alone, although in some mice (6/10) tumors did
regrow after treatment was discontinued. Similar synergy was observed with lenalidomide plus
anti-CD38-IFNċ(att) combination treatment of mice bearing lenalidomide-refractorytumors
(data not shown). These data suggest that even MM tumors refractory to treatment with a pro-
teasome inhibitor can respond to combined treatment with a targeted, attenuated IFNċ.
Discussion
Our experiments confirm previous reports that IFNċhas anti-myeloma activity by showing
that every primary MM and MM cell line we tested was sensitive to its growth inhibitory effects
at high doses. Despite strong anti-tumor activity, however, clinical use of IFNċis limited due
to significant side effects at therapeutic doses. Here, we describe a novel immunological
approach to minimize off-target IFNċtoxicity while retaining robust on-target anti-tumor
activity. This was accomplished by engine ering an attenuating mutation into the IFNċportion
of an immunoc ytokine targeted to CD38 on MM tumor cells.
Multiple groups have shown that targeting IFNċto tumor cells improves anti-tumor activ-
ity in animal studies [16,19,20,24,47]. Treatment with CD20-targeted mouse IFNċ,for
example, inhibited B-cell lymphoma tumor growth resulting in an 87% tumor cure rate in mice
[19]. Similarly, Rossi et al. demonstrated strong anti-tumor activity of a CD20-targeted tetra-
meric IFNċin xenograft NHL models, and Yoo et al. showed delayed tumor growth with a
CD138-targeted IFNċin a MM xenograft model [24,47]. While these experiments convinc-
ingly demonstrate the utility of targeted IFNċimmunocytokines against tumors they do not
address the well documented issue of IFNċtoxicity. These targeted approaches with wild type
cytokines may, in fact, produce significant toxicities given the prolonged half-life of immuno-
cytokines compared to free cytokines.
Garcin et al. recently reported use of an attenuated form of IFNċto reduce cell toxicity [21].
They fused a mutated mouse IFNċto a mLepR targeting nanobody and demonstrated strong
reporter activity on target-positive HL116 cells with negligibleactivity on target-negative cells.
Importantly, the most effective dose at inducing strong STAT1 phosphorylation in vitro had
no ef fect on target-negative cells in vivo. We expanded upon t his strategy by specifically target-
ing attenuated human IFNċto MM tumor cells via an anti-CD38 antibody and then tested its
anti-tumor activity in vivo.
The CD38-targeted attenuated IFNċfusion protein, or “Attenukine™”, identified as anti-
CD38-IFNċ(att), displayed a 10,000-fold greater specificity than native IFNċfor CD38-posi-
tive (tumor) vs CD38-negative (normal) cells (Fig 3C and 3D). In contrast, the corresponding
wild type IFNċfusion protein showed only a 40-fold greater specificity (Fig 3A and 3B). There-
fore, the attenuating mutation in the IFNċportion of the immunocytokine increased the anti-
gen-specificity of IFNċby approximately 250-fold. This suggests that patients might safely be
treated with higher doses of the anti-CD38-IFNċAttenukine™ compared to native IFNċor a
conventional IFNċimmunocytokine.
Other experiments reported here support a potentially favorable safety profile for an IFNċ
Attenukine™. Anti-CD38-IFNċ(att) inhibition of normal human BM cell colony growth in
vitro was ~6,000 fold weaker than native IFNċ. Additionally, in a pilot cynomolgus monkey
CD38-Targeted , Attenuated IFNĮIs Potent with Decreased Off-Target Activity
PLOS ONE | DOI:10.1371/journal.pone.0162472 September 9, 2016 16 / 20
study, a non-targeted attenuated IFNċfusion protein was more than a 100-fold weaker at
inducing systemic pro-inflammatory markers than the corresponding wild type IFNċfusion
protein confirming that the attenuating mutation in IFNċdoes indeed result in lower off-target
IFNċactivity.
Despite strong attenuation, the CD38-targeted, attenuated IFNċprovided dramatic MM
anti-tumor activity in various cell line and primary-derived tumor models. In fac t, anti-
CD38-IFNċ(att) completely eliminated established NCI-H929 MM subcutaneous tumors in all
mice, and fully resolved even very large tumors. We are not aware of any other compound that
has shown such profound preclinical anti-myeloma activity in vivo, especially on large estab-
lished tumors. In comparison to various approved MM drugs, anti-CD38-IFNċ(att) was signif-
icantly more effective in the NCI-H929 xenograft model. In addition, the strong synergistic
effects of anti-CD38-IFNċ(att) with both bortezomib and lenalidomide indicates that drug
combinations may offer an optimal regimen in the clinic, even for bortezomib-refractory MM
tumors. Finally, Daudi lymphoma tumors were highly responsive to IFNċAttenukine™ indicat-
ing that anti-CD38-IFNċ(att) is likely a strong therapeutic candidate for multiple types of
CD38-positive lymphoid cancers. It is important to note that while it would be most ideal to
test both therapeutic activity and tolerability of IFNċand related constructs in the same animal
model, this is not currently possible. Murine cells do not respond significantly to human IFNċ,
making mice inappropriate models for IFNċtoxicity experiments. Cynomolgus monkeys serve
as good animal models for tolerability to IFNċbecause monkey cells respond to human IFNċ,
but testing efficac y of drug compounds in monkeys is not possible or desirable. Thus, here we
present tumor inhibition studies performed solely in mice and preliminary IFNċtolerability
studies performed in monkeys.
A number of monoclonal antibodies targeting MM antigens such as CD38 and CS1
(SLAMF7) have entered clinical development [48]. These antibodies lead to immune effector-
mediated killing and may eliminate both normal and tumor cells expressing target antigen.
Based on our results, the IgG4 anti-CD38 antibody fused to IFNċin our constructs would not
be expected to kill CD38
+
normal cells by antibody-dependent cell-mediated cytotoxicity
(ADCC) or complement-dependent cytotoxicity (CDC). Rather, only IFNċ-sensitive, CD38
high-expressing cells, such as MM cells, will be targeted and killed by this more selective thera-
peutic approach.
In summary, we describe a novel, targeted IFNċAttenukine™ that harnesses the potent anti-
tumor activity of native IFNċwhile reducing its well documented toxic side effects. The potent
anti- tumor activity with predicted lower toxicity together may provide a broader therapeutic
index (TI) for the IFNċAttenukine™ compared to native IFNċor conventional IFNċimmuno-
cytokines. The remarkably durable anti-tumor activity of this new compound is unsurpassed
by any current small or large molecules tested in similar MM models. Studies are underway to
elucidate the possible mechanisms of action of CD38-targeted IFNċAttenukine™.
Acknowledgments
We thank Darlene Jenkins, Ph.D., a writer funded by Teva, for providing drafts and edits of
this manuscript. We gratefully acknowledge Maxwell Stevens and Teresa Domagala for provi-
sion of reagents and Simon Tout for thorough review of this manuscript.
Author Contributions
Conceptualization: SP DW SN.
Investigation: SP TT MB YY AS GM CB MS HH ES HC.
CD38-Targeted , Attenuated IFNĮIs Potent with Decreased Off-Target Activity
PLOS ONE | DOI:10.1371/journal.pone.0162472 September 9, 2016 17 / 20
Metho dolog y: SP DW TT MR MS JB HC ES.
Resources: TT YY AS JB.
Superv ision: SP DW.
Writing – original draft: SP DW.
Writing – revie w & editing: SP AD DW MR CB MS JB.
References
1. Institute NC. SEER stat fact sheets: myeloma [cited 2015 September 10]. Available from: http://seer.
cancer.gov/statfacts/html/mulmy.html.
2. Raab MS, Podar K, BreitkreutzI, Richardson PG,Anderson KC. Multiple myeloma. Lancet. 2009; 374
(9686):324–39. doi: 10.1016/S0140-6736(09)60221-X PMID: 19541364.
3. El-Amm J, Tabbara IA. Emerging Therapies in Multiple Myeloma. Am J Clin Oncol. 2013. doi: 10.1097/
COC.0b013e3182a4676b PMID: 23934133.
4. Siegel R, Ma J, Zou Z, Jemal A. Cancer statistics, 2014. CA Cancer J Clin. 2014; 64(1):9–29. doi: 10.
3322/caac.21208 PMID: 24399786.
5. Belardelli F, Ferrantini M, Proietti E, Kirkwood JM. Interferon-alpha in tumor immunity and immunother-
apy. Cytokine Growth Factor Rev. 2002; 13(2):119–34. PMID: 11900988.
6. Rizza P, Moretti F, BelardelliF. Recent advances on the immunomodulatory effects of IFN-alpha: impli-
cations for cancer immunotherapy and autoimmunity. Autoimmunity. 2010; 43(3):204–9. doi: 10.3109/
08916930903510880 PMID: 20187707.
7. Mellstedt H, Ahre A, Bjorkholm M, HolmG, Johansson B, Strander H. Interferon therapy in myelomato-
sis. Lancet. 1979; 1(8110):245–7. PMID: 84901.
8. Fritz E, Ludwig H. Interferon-alpha treatment in multiple myeloma: meta-analysis of 30 randomised tri-
als among 3948 patients. Ann Oncol. 2000; 11(11):1427–36. PMID: 11142483.
9. Al-Zahrani H, Gupta V, Minden MD, Messner HA, Lipton JH. Vascular events associated with alpha
interferon therapy. Leuk Lymphoma. 2003; 44(3):471–5. doi: 10.1080/1042819021000055066 PMID:
12688317.
10. Kirkwood J. Cancer immunotherapy: the interferon-alpha experience. Semin Oncol. 2002; 29(3 Suppl
7):18–26. PMID: 12068384.
11. Ludwig H, Cohen AM, Polliack A, Huber H, Nachbaur D, Senn HJ, et al. Interferon-alpha for induction
and maintenance in multiple myeloma: results of two multicenter randomized trials and summary of
other studies. Ann Oncol. 1995; 6(5):467–76. PMID: 7669712.
12. O'Brien S, Kantarjian H, Talpaz M. Practical guidelines for the management of chronic myelogenous
leukemia with interferon alpha. Leuk Lymphoma. 1996; 23(3–4):247–52. doi: 10.3109/
10428199609054827 PMID: 9031105.
13. Schaar CG, Kluin-Nelemans HC, Te Marvelde C, le Cessie S, Breed WP, Fibbe WE, et al. Interferon-
alpha as maintenance therapy in patients with multiple myeloma. Ann Oncol.2005; 16(4):634–9. doi:
10.1093/annonc/mdi125PMID: 15741226.
14. List T, Neri D. Immunocytokines: a review of molecules in clinical development for cancer therapy. Clin
Pharmacol. 2013; 5:29–45. doi: 10.2147/CPAA.S4923 1 PMID: 23990735; PubMed Central PMCID:
PMC3753206.
15. Pasche N, Neri D. Immunocytokines: a novel class of potent armed antibodies. Drug Discov Today.
2012; 17(11–12):583–90. doi: 10.1016/j.drudis.2012.01.007 PMID: 22289353.
16. Huang TH, Chintalacharuvu KR, Morrison SL. Targeting IFN-alpha to B cell lymphoma by a tumor-spe-
cific antibody elicits potent antitumor activities. J Immunol. 2007; 179(10):6881–8. PMID: 17982079.
17. Ozzello L, De Rosa CM, Blank EW, Cantell K, Ceriani RL, Habif DV Sr. The use of natural interferon
alpha conjugated to a monoclonal antibody anti mammary epithelial mucin (Mc5) for the treatment of
human breast cancer xenografts. Breast Cancer Res Treat. 1993; 25(3):265–76. PMID: 8369527.
18. Rossi EA, Rossi DL, Cardillo TM, Stein R, Goldenberg DM, Chang CH. Preclinical studies on targeted
delivery of multiple IFNalpha2b to HLA-DR in diverse hematologic cancers. Blood. 2011; 118(7):1877–
84. doi: 10.1182/blood-2011-03-343145 PMID: 21680794; PubMed Central PMCID: PMC3158718.
19. Xuan C, Steward KK, Timmer manJM, Morrison SL. Targeted delivery of interferon-alpha via fusion to
anti-CD20 results in potent antitumor activity against B-cell lymphoma. Blood. 2010; 115(14):2864–71.
doi: 10.1182/blood-2009-10-250555 PMID: 20139095; PubMed Central PMCID: PMC2854431.
CD38-Targeted , Attenuated IFNĮIs Potent with Decreased Off-Target Activity
PLOS ONE | DOI:10.1371/journal.pone.0162472 September 9, 2016 18 / 20
20. Frey K, Zivanovic A, Schwager K, NeriD. Antibody-based targeting of interferon-alpha to the tumor
neovasculature: a critical evaluation. Integr Biol (Camb). 2011; 3(4):468–78. doi: 10.1039/c0ib00099j
PMID: 21225052.
21. Garcin G, Paul F, Staufenbiel M, Bordat Y, Van der Heyden J, Wilmes S, et al. High efficiency cell-spe-
cific targeting of cytokine activity. Nat Commun. 2014; 5:3016. doi: 10.1038/ncomms4016 PMID:
24398568.
22. Thakur ML, Li J, Donegan M, Pallela VR, Kolan H, Duggaraju R, et al. Improved antibody targeting with
interferon-alpha-2bconjugate. J Immunother. 1997; 20(3):194–201. PMID: 9181457.
23. Yang X, Zhang X, Fu ML, WeichselbaumRR, Gajewski TF, Guo Y, et al. Targeting the tumor microenvi-
ronment with interferon-beta bridges innate and adaptive immune responses. Cancer Cell. 2014; 25
(1):37–48. doi: 10.1016/j.ccr.2013.12.004 PMID:24434209; PubMed Central PMCID: PMC3927846.
24. Yoo EM, Trinh KR, Tran D, Vasuthasawat A, Zhang J, Hoang B, et al. Anti-CD138-targeted interferon is
a potent therapeutic against multiple myeloma. J Interferon Cytokine Res. 2015; 35(4):281–91. doi: 10.
1089/jir.2014.0125 PMID: 25353626; PubMed Central PMCID: PMC4390003.
25. Kim D, Park CY, Medeiros BC, Weissman IL. CD19-CD45 low/- CD38 high/CD138+ plasma cells
enrich for human tumorigenic myeloma cells. Leukemia. 2012; 26(12):2530–7. doi: 10.1038/leu.2012.
140 PMID: 22733078.
26. Lin P, Owens R, Tricot G, Wilson CS. Flow cytometricimmunophenotypic analysis of 306 cases of mul-
tiple myeloma. Am J Clin Pathol. 2004; 121(4):482–8. doi: 10.1309/74R4-TB90-BUWH-27JX PMID:
15080299.
27. Santonocito AM, Consoli U, Bagnato S, Milone G, Palumbo GA, Di Raimondo F, et al. Flow cytometric
detection of aneuploid CD38(++)plasmacells and CD19(+) B-lymphocytes in bonemarrow, peripheral
blood and PBSC har vestin multiple myeloma patients. Leuk Res. 2004; 28(5):469–77. doi: 10.1016/j.
leukres.2003.09.015 PMID: 15068900.
28. de Weers M, Tai YT, van der Veer MS, Bakker JM, Vink T, Jacobs DC, et al. Daratumumab, a novel
therapeutic human CD38 monoclonal antibody, induces killing of multiple myeloma and other hemato-
logical tumors. J Immunol. 2011; 186(3):1840–8. doi: 10.4049/jimmunol.1003032 PMID: 21187443.
29. Schlesinger JJ, Brandriss MW, Monath TP. Monoclonal antibodies distinguish between wild and vac-
cine strains of yellow fever vir us by neutralization, hemagglutination inhibition, and immune precipita-
tion of the virus envelope protein. Virology. 1983; 125(1):8–17. PMID: 6187129.
30. Piehler J, Roisman LC, Schreiber G. New structural and functional aspects of the type I interferon-
receptor interaction revealed by comprehensive mutational analysis of the binding interface. J Biol
Chem. 2000; 275(51):40425–33. doi: 10.1074/jbc.M006854200 PMID: 10984492.
31. Durocher Y, Perret S, Kamen A. High-level and high-throughput recombinant protein production by
transient transfection of suspension-growing human 293-EBNA1 cells. Nucleic Acids Res. 2002; 30(2):
E9. PMID: 11788735; PubMed Central PMCID: PMC99848.
32. Committee NRCU. Guide for the care and use of laboratory animals, 8th edition. Washingon (DC):
National Academies Press (US); 2011.
33. Olfert ED, Cross BM, McWilliam AA. Guide to the care and use of experimental animals, 2nd edition:
Canadian Council on Animal Care; 1993.
34. Dryden MW, Payne PA, Ridley R, Smith V. Comparison of common fecal flotation techniques for the
recovery of parasite eggs and oocysts. Vet Ther. 2005; 6(1):15–28. PMID: 15906267.
35. Campbell RA, Manyak SJ, Yang HH, Sjak-Shie NN, Chen H, Gui D, et al. LAGlambda-1: a clinically rel-
evant drug resistant human multiple myeloma tumor murine model that enables rapid evaluation of
treatments for multiple myeloma. Int J Oncol. 2006; 28(6):1409–17. PMID: 16685443.
36. Campbell RA, Berenson JR. Animalmodels of multiplemyeloma and theirutility in drug discovery. Curr
Protoc Pharmacol. 2008;Chapter 14:Unit 14 9. doi: 10.1002/0471141755.ph1409s40 PMID:
22294221.
37. Friedman RM. Clinical uses of interferons. Br J Clin Pharmacol. 2008; 65(2):158–62. doi: 10.1111/j.
1365-2125.2007.03055.x PMID: 18070219; PubMed Central PMCID: PMC2253698.
38. Trown PW, Wills RJ, Kamm JJ. The preclinical development of Roferon-A. Cancer. 1986; 57(8
Suppl):1648–56. PMID: 3081245.
39. Stewart WEI. "Species specificity" of interferons. The Interferon System. Wein, New York: Spr inger-
Verlag; 1981. p. 135–45.
40. Chillemi A, Zaccarello G, Quarona V, Lazzaretti M, Martella E, Giuliani N, et al. CD38 and bone marrow
microenvironment. Front Biosci (Landmark Ed). 2014; 19:152–62. PMID: 24389178.
41. Osborn BL, Olsen HS, Nardelli B, Murray JH, Zhou JX, Garcia A, et al. Pharmacokinetic and pharmaco-
dynamic studies of a human serum albumin-interferon-alpha fusion protein in cynomolgus monkeys. J
Pharmacol Exp Ther. 2002; 303(2):540–8. doi: 10.1124/jpet.102.037002 PMID: 12388634.
CD38-Targeted , Attenuated IFNĮIs Potent with Decreased Off-Target Activity
PLOS ONE | DOI:10.1371/journal.pone.0162472 September 9, 2016 19 / 20
42. Jeon S, Juhn JH, Han S, Lee J, Hong T, Paek J, et al. Saturable human neopterin response to inter-
feron-alpha assessed by a phar macokinetic-pharmacodynamic model. J Transl Med. 2013; 11:240.
doi: 10.1186/1479-5876-11-240 PMID: 24088361; PubMed Central PMCID: PMC3853247.
43. Padovan E, Spagnoli GC, Ferrantini M, Heberer M.IFN-alpha2ainduces IP-10/CXCL10 and MIG/
CXCL9 production in monocyte-deriveddendriticcells and enhances their capacityto attract and stimu-
late CD8+ effector T cells. J Leukoc Biol. 2002; 71(4):669–76. PMID: 11927654.
44. Boccadoro M, MorganG, Cavenagh J. Preclinical evaluation of the proteasomeinhibitor bortezomib in
cancer therapy. Cancer Cell Int. 2005; 5(1):18. doi: 10.1186/1475-2867-5-18 PMID:15929791;
PubMed Central PMCID: PMCPMC1164423.
45. Rozewski DM, Herman SE, Towns WH 2nd, Mahoney E, Stefanovski MR, Shin JD, et al. Pharmacoki-
netics and tissue disposition of lenalidomide in mice. AAPS J. 2012; 14(4):872–82. doi: 10.1208/
s12248-012-9401-2 PMID: 22956478; PubMed Central PMCID: PMCPMC3475844.
46. McLaughlin F, Mackintosh J, Hayes BP, McLaren A, Uings IJ, Salmon P, et al. Glucocor ticoid-induced
osteopenia in the mouse as assessed by histomor phometr y, microcomputed tomography, and bio-
chemical markers. Bone. 2002; 30(6):924–30. PMID: 12052464.
47. Rossi EA, Goldenberg DM, Cardillo TM, Stein R, Chang CH. CD20-targeted tetrameric interferon-
alpha, a novel and potent immunocytokine for the therapy of B-cell lymphomas. Blood. 2009; 114
(18):3864–71. doi: 10.1182/blood-2009-06-228890 PMID: 19710501; PubMed Central PMCID:
PMC2773491.
48. Sondergeld P, van de Dong NWCJ, Richardson PG, Plesner T. Monoclonal antibodies in myeloma.
Clin Adv Hematol Oncol. 2015; 13(9):599–609. PMID: 26452191
CD38-Targeted , Attenuated IFNĮIs Potent with Decreased Off-Target Activity
PLOS ONE | DOI:10.1371/journal.pone.0162472 September 9, 2016 20 / 20
